The COMIS Plan: IT-Mediated Business Reengineering in Telecom Australia
during the 1960s
Doug Hamilton Martin Atchison
Department of Information Systems Monash University, Caulfield Campus
Abstract
The authors suggest in this paper that the understanding of business reengineering as a current discipline can be enhanced by research into past process redesign initiatives which appear to be based on similar assumptions, particularly in relation to the role of IT as the essential enabler of change. AI> an instance of this approach, the paper deals with a business reenginee1ing plan developed by Telecom Australia during the 1960s. After several years of implementation activities, the plan was abandoned in 1972. The authors conclude that the assumptions made in preparing the plan effectively precluded adequate consideration of the practical issues involved. They suggest that business enterprises undertaking major reengineering initiatives in the 1990s may be at risk of underestimating the difficulties involved for essentially the same reason.
1. INTRODUCTION
Business reengineering has been widely canvassed as a means of achieving radical change in business structures and processes. Some dramatic claims have been made for the power of reengineering concepts. Hammer and Champy ( 1993) state that "most companies have no choice but to muster the courage to do it [reengineer]. For many, reengineering is the only hope for breaking away from the ineffective, antiquated ways of conducting business that will otherwise inevitably destroy them." In their view, business processes are currently institutionalized in fragmented, disjunct forms that are inimical to the effective management of business as a whole. Their recommended alternative is a business redesign based on the reintegration of underlying logical processes, enabling a better focus on customer se1vice and business performance.
90 Part Two Images of Practice: Business Process Reengineering
Much of the reengincering literature is extremely positive in tone. A leading theorist comments
that
in some companies, I have encountered situations where managers invoked t1imsy
and unsubstantiated excuses- "we tried something like this and it didn't work"
or "we can't afford to make such changes now" - to prevent initiation of the
higher levels of transformation. [Venkatraman 1994]
We take the management comments in this quotation to represent legitimate practical concerns,
however. Research into a related area of IT application shows that a longer-term perspective can
cast doubt on the validity of strong claims made for the benefits of assigning IT a central role in
the development of corporate strategies. In a study of "classic cases" of strategic information
systems, Kettinger, Grover and Segars ( 1995) found that the business advantages identitied as
deriving from the successful application of IT to achieve strategic business goals could not in all
cases be confirmed by objective performance measures. Merrill Lynch, CIGNA and Citicorp
were, for example, among a number of "high-profile" case study firms that showed reduced
profitability relative to industry averages over an extended period following the implementation
of a strategic system. The authors concluded that "too often, decisions concerning the
introduction of strategic IT have been based on management's ability to make a leap of faith"
(Kettinger, Grover and Segars 1995). We infer that a company investing time and resources in
applying business reengineering concepts may well be making just such a leap of faith.
The problem for empirical research into the business value of reengineering in its current form is
that there is as yet a shortage of case material from which to develop a long-term perspective on
the full practical implications of applying the concepts. We suggest that one way of supplement
ing research into examples of reengincering as a contemporary phenomenon is to investigate
earlier approaches to IT-mediated organizational change deemed to embed analogous principles.
To thi~ end, we examine the aims and effects of a major corporate IS plan (known as CO MIS
Computer Operations and Management Information System) developed during the 1960s in
Telecom Australia. Comprehensive in its scope, the plan was intended to provide a framework
for the development of an integrated structure of programs and files able to be reconfigured as
required to support changing organizational process structures. Activities to implement the plan
were undertaken and pursued for some years, but abandoned when a number of major practical
difficulties were encountered.
The suggestion that a parallel can be drawn between COMIS and contemporary business
reengineering theory is based on our view that the key concepts underpinning reengineering are
a set of assumptions analogous to those deemed to provide the basis for the development of the
COMIS plan. Broadly speaking, these assumptions are taken to be, first, that information
technology has evolved to a level of power and sophistication sufficient to alter the fundamental
logic of business process design and, second, that existing business processes arc inefficient
enough to warrant redesign. On the basis of this perspective we suggest that the ultimate failure
of CO MIS has implications relevant to the further development of reengincering theory.
IT-mediated business reengineering in Telecom Australia 91
2. BUSINESS REENGINEERING THEORY
Based on a review of four years' experience with the application of business reengineering principles in a number of companies, Davenport and Stoddard (1994) identify five concepts as constituting the basic theory:
I. A clean slate approach to organizational design and change. 2. An orientation to broad, cross-functional business processes, or how work is done. 3. The need for, and possibility of, radical change in process performance. 4. Information technology as an enabler of change in how work is done. 5. Changes in organi7.ational and human arrangements that accompany change in
technology.
Davenport and Stoddard suggest that these concepts can all be found in previous theories of organizational change, and that business reengineering is new only to the extent that wellestablished concepts have been "combined in a new synthesis." While other theorists make stronger claims for the novelty of reengineering as an approach to organi7.ational change (Hammer 1990; Hammer and Champy 1993; Venkatraman 1991, 1994), the synthesis of a process orientation with the concept of modem IT as the enabler of change is the common thread running through different representations of the approach.
Implicit in the view of IT as the "essential enabler of process change" (Hammer and Champy 1993) is the assumption that it is IT which has changed the logic of doing business sufficiently to generate a need for fundamental change. In general the nature of this assumption is not explored in detail. Hammer and Champy do not, for instance, indicate whether in their opinion the possibility of radical change through business process redesign has been inherent in the business environment since the first emergence of powerful new information technologies such as telegraphy, or alternatively whether the continuing evolution in both computing and telecommunications technologies has resulted in a qualitative change in the ways in which the technology can be applied.
Yates and Benjamin (1991) provide a specific rationale at least for a claim that there has been a qualitative change in the power of IT. They identify the four key functions of IT as information conversion, information storage, information processing and information communications, and state that in their view a major turning point has been reached with the possibilities for the integration of communications and data processing capabilities raised by the convergence of computing and telecommunications technologies. The example quoted is one in which the four key functions can be integrated in relation to a specitic business activity at one workstation. Venkatraman cites this research as support for the view that IT "can be thought of as a new engine for the organization" and that IT is now at a point where it renders "some modes of organizing relatively inefficient" (Venkatraman 1991).
We take the assumption that modem IT has effectively changed some of the rules for doing business to be extremely strong. Proponents of reengineering do not simply suggest that the organizational structures supporting business processes should be redesigned to accommodate the incorporation of IT in those processes, but claim instead that the logical basis for redesign has
92 Pan Two Images of Practice: Business Process Reengineering
changed (Hammer and Champy 1993, Venkatraman 1994). In our view this assumption is implicitly based on an abstract view of IT as a "given" in the business environment, or in other words on the idea that IT considerations can for the purposes of redesign be considered as a separate issue. We discuss later whether adopting this view might have a tendency to mislead planners and designers as to the true complexity of process issues in a dynamic business environment, where the logic of existing structures is continually being challenged by changes in the competitive situation. At this point however, it is relevant to point out that the technology itself remains problematic to some extent; in common with Keen (1991), Yates and Benjamin strike a strong note of caution, pointing out that integration techniques are still quite immature and face a number of compatibility problems. Keen makes the additional point that firms face "many technical uncertainties as they try to balance conflicting demands for efficiency now and integration tomorrow."
To provide a background to our discussion of the case, we suggest that the view of integrated IT as an "essential enabler of change" is strongly paralleled by views on the promise of computing for business during the 1960s. In the first flush of enthusiasm for computing, there seemed no real limit to its potential A leading theorist of the time commented in 1960 that "within the very near future - much less than twenty-five years- we will have the technological capability of substituting machines for any and all human functions in organizations" (Simon 1970). Weizenbaum notes that the computer was viewed as "the ultimate abstract tool" and that a view emerged in the years after World War II that it had providentially arrived to rescue industry from problems caused by the failure of existing organizational systems to keep up with the increasing pace of business:
managers and technicians agreed that the computer had come along just in time to avert catastrophic crises; were it not for the timely introduction of computers, it was argued, not enough people could have been found to staff the banks, the ever increasingly complex communication and logistic problems of armed forces spread all over the world could not have been met, and trading on the stock and commodity exchanges could not have been maintained. [Weizenbaum 1976]
3. METHODOLOGY
The case discussion is based on qualitative data from a study of the information systems (IS) function in Telecom Australia (referred to in the rest of this paper simply as Telecom) during the period 1960-1995. The CO MIS plan discussed in this paper emerged during the course of the study as an activity that played a significant role in shaping future IS directions in the company, and was consequently made a specific study focus. The primary source documentation available on COMIS was extensive and was the major source of data. Supporting data was obtained from the five (of more than forty) people interviewed in the course of the total study who had direct experience ofCOMIS. Part of the interview in each of these five cases was devoted specifically to CO MIS and its repercussions.
A constructivist approach to the study was adopted. The constructivist paradigm implies that it is the researcher's job to provide sufficient contextual information in the form of "thick description" (Geertz 1973) for a reader to make an individual assessment of the relevance of the
IT-mediated business reenginecring in Telecom Australia 93
findings within the reader's frame of reference (Firestone 1990). Thus while a number of parallels between COMIS concepts and those underpinning business reengineering became apparent during the course of the analysis, the changes in business, technology and tenninology that have occurred since 1970 were deemed sufficient to render direct comparisons and evaluations redundant. The analytical method used instead was to examine the data for evidence of prejudgments and expressed intents which appeared to be strongly analogous to those being presented in the context of business reengineering, to analyze the findings in the light of IT experience over the subsequent years, and to identify what in our view are the implications for reengineeling theory.
4. THE COMIS PLAN IN TELECOM AUSTRALIA 1960-1972
During the period from 1960 to 1972, Telecom Australia (Telecom) was responsible for administering a national monopoly in the telecommunications industry. Charged with a number of social responsibilities, including the development of the telephone network as a key national asset, the organiz.ation had a strong vested interest in the fully planned and managed introduction of new technology.
There was a strong technological strain in the Telecom culture during that time. The key corporate values were derived from engineering considerations of efficiency and quality of workmanship, and a belief in the efficacy of technology as a solution for social as well as technical problems. Moyal (1984) records that Telecom staff were in general well aware of the social roles played by telecommunications in a sparsely populated country such as Australia, and that there was considerable corporate pride in the technical achievements that had been required to build the cable networks linking remote rural areas with the rest of Australia, and Australia with the rest of the world. The result was an organizational environment extremely receptive to the introduction of new technology.
A number of small computer applications were developed in Telecom during the late 1950s. These were deemed by management to have been sufficiently successful to warrant an in-depth exploration of the possibilities of the new technology, and a major planning study was initiated in 1959. In its final form, the COMIS plan for the development of a set of integrated applications was the result of this initial study, complemented by a further study duling 1967 and 1968.
4.1 The ADP Exploratory Study 1959-1960
The Telecom documentation indicates that the application of computers to business processes was already being considered in strategic terms late in the 1950s. The report from the 1959 study group stated that "the Post Office is ready for A.D.P. and .. .it.~ introduction is as necessary and inevitable as was the introduction of automatic telephony fifty years ago. It only remains for us to plan to introduce the best system as efficiently and speedily as possible" (Australian Post Office 1961 ). The concept of an integrated system which would automate many of Telecom's business processes was introduced at this time, when it was thought it would handle all of the "business processes susceptible to electronic data processing" (Australian Post Office 1959).
There is little indication in the study group report of serious consideration having been given to the full range of organiz.ational impacts contingent on the widespread introduction of computing
94 Part Two Images of Practice: Business Process Reengineering
technology. There was instead a tendency throughout the report to identify the proposed applications with the technology, as evidenced in a discussion of the benefits expected to be achieved. These were listed as:
(I) The ability of the equipment to hold very large volumes of records at a central point; (2) The ability of the equipment to extract from these records, with practically no
expenditure of human effort, information in any desired form; (3) The ability of the equipment to make logical decisions in conformity with predeter
mined rules; (4) The ability of the equipment to print out information based on its records without
significant human effort; (5) The fact that the equipment rarely makes mistakes. [Australian Post Office 1961]
The planning orientation was at this stage largely toward the automation of existing processes.
Interviewees put forth the view that Telecom planners were just commencing to consider the
possibilities of the technology in more radical terms. This can be detected in the study group
report as a view that systems integration could enable the organization to rationalize clerical activities in specific areas.
In order to obtain maximum use from the input data, and to obtain the benefits of an integrated system, the central computer should process all the work within a specific geographical area for which the system as a whole is designed. The use of one computer [i.e., application or system] to handle, say, telephone accounts, another to handle payroll work and so on would not be as economical or efficient as the integrated system envisaged in this report. It is probably true that such a course of action would permit the introduction of A.D.P. on a limited scale earlier than would otherwise be the case. [Australian Post Office 1961]
The initial scope of the system was specified in this study. Work commenced on the first
application (the Telephone Accounting or TEL/DRS system) in 1963, which was released into
production in suburban Melbourne in 1967. In 1968, the term CO MIS (Computer Operations
and Management Information System) was introduced to refer to the integrated system, and its scope was considerably expanded in another major study.
4.2 The Study Group on Application of Computers 1967-1968
The Telecom IS area had gained considerable experience in the construction of computer
applications for business by 1967, when the second major systems study was commenced. The
TEL/D.S. application already in production was a large system, with more than $3M invested in
it by 1966 (Australian Post Office 1966). The study group report was much more sophisticated
in its analysis of IS issues than its predecessor, with a specific focus on the capabilities for
computing to facilitate radical change. The two major emphases were, first, on the potential for
computing to enable the organization to stay in control of evolving business requirements and,
second, on a perceived opportunity to develop a system that could be reconfigured as required
to meet changing organizational needs. The approach was, however, not based specifically on
IT-mediated business reengineering in Telecom Australia 95
theoretical considerations and justification for the proposal was couched in terms of quantitied capital and operational costs savings rather than more abstract strategic considerations.
A significant aspect to the expanded CO MIS plan was the conceptualization of Telecom's set of business processes as an engineered structure forming a network in principle little different from any other network. The inference drawn was that the application of IT to business processes could be treated as primarily an abstract engineering design issue. The planners viewed the process structure as "wide-ranging and complex; it is comprehensive in its way though heavily compartmentalized and limited by the means it employs" (Australian Post Office 1968).
In addressing the issue of change, the report noted that
telecommunications is a key activity for all economic and social progress, requiring capital investment at a rate that greatly exceeds the economic growthrate of the community as a whole .... There is often a temptation to believe that some sort of a plateau of expansion is being reached; subsequent events invariably disprove this. [Australian Post Office 1968]
The report went on to an analysis of the implications of the rates of change then being experienced and concluded that an integrated system was required to deal with the likely effects. For the time, the system envisaged was huge:
the complete CO MIS will be a vast system catering for the information needs of practically all Departmental activities. It will include on-line storage of some 20,000 million characters, high capacity processors with security back-up and a communications network giving access to some 5000 Departmental stations including terminal equipment ranging from teleprinters to visual displays. The capital cost of such a system would be around $90M .... System design and implementation would amount to some 1500 man-years. Nevertheless the system would be highly profitable and would, in fact, represent the only feasible solution to the immense management problems of scale and complexity that will confront the Department in the future. [Australian Post Oftice 1968]
The planners addressed a number of data issues in the development of COMIS. While the comments were somewhat scattered, taken as a whole they were sufficiently comprehensive to form the basis for what would be a data management policy in more recent te1minology. Recommendations included the standardization of codes and terminology, the once-only rendering of information in permanent store, and the high-priority restructuring of the telephone service occupancy tile so that it could support the full range of business processes requiring occupancy data.
The integrated system was no longer seen simply as an end in itself, but as a means of restructuring the organization's business processes along more t1exible lines. An interviewee working in a business unit at the time said that "we were fmally getting round to some idea of commercialization and the customer. It was very early days though and most of the engineers
96 Part Two 111Ulges of Practice: Business Process Reengineering
wouldn't have a bar of it." The implications of a more commercial approach were, however, recognized by the study group:
where the increase [in information requirements] is purely in volume, it may be practicable to meet it by corresponding expansion of existing methods. However, more typically there is an increase in complexity ... .Increasing emphasis on the Department as a business enterprise is generating demands for information needed
to exercise new controls .... For these new methods must be found. [Australian Post Office 1968]
The solution proposed was an almost seamless collection of sharable files and processes. The
fundamental concept was that the structure of files and programs would be designed so that it could be dynamically reconfigured to form new applications as needs changed. To a very
considerable extent, the study group was constrained to speak in terms of tixed applications and functions for the report to be generally intelligible within Telecom, but the report notes that where applications were mentioned this was "only a means of study and reporting" and that the referents
were not to be taken to be self-contained systems. Overall there was a clear process orientation
to the plan. The stated intention was to do away with the "compartmentalization" of manual
processes mentioned earlier and to use IT to establish a highly flexible structure. The long-term objective for the IS area was to be
a comprehensive COMIS, which would provide a framework within which all future applications should be considered. In this development, program and tile strategies as well as codes and terminology should take account not only of contemporary applications but also needs envisaged for the future. [Australian Post Office 1968]
An interviewee who was subsequently a senior IS executive for many years commented that "CO MIS was more about reengineering the business than anything I've seen since. Some of the people on COMIS were quite radical and would have gone further but felt you couldn't get away
with changing too much at once."
It is noteworthy that the planning report deals with the issues involved almost exclusively in
technical terms. The report does not contain any form of 1isk analysis, nor any significant
reference to potential problems in relation to the management of separate but logically dependent
projects. The report conveys no sense at all that the development of CO MIS was expected to be
anything more than a straightforward technical exercise.
4.3 The COMIS Review 1971
With the TEUDRS system discussed earlier providing a foundation, the first major step in
realizing the "ultimate system" (Australian Post Office 1968) was to be the development of an
application to automate telephone directory compilation in conjunction with a redevelopment of
TEL/DRS to form an integrated suite of programs and files. A number of problems were
encountered and in 1971 a major review of COMIS activities resulted in the conclusion that
"although considerable benefits in accuracy, control and economy could be obtained through a
IT-mediated business reengineering in Telecom Australia 97
total integrated system, the conclusion has been reached that it is probably beyond the resources, experience, management and technical capability of the Department [to achieve] at this stage" (Australian Post Office 1971 ).
The review team identified a number of practical problems. Most significant was that of dealing simultaneously with the difficulties of building two large systems, as well as managing the interactions between them. It was commented that
the extended system development periods associated with the comprehensive nature and considerable resource requirements for system testing were not previously appreciated and are seen as major difficulties. It must be appreciated also that large scale systems tend to be fairly rigid and do not permit changes to be made easily for changed user needs either during development or subsequently. [Australian Post Office 1971]
A related issue was a "snowball" effect on resource and management overhead requirements as the business environment continued to evolve and new requirements for systems support emerged. The review report notes that
since the original CO MIS study was completed, computer controlled telecommunications equipment has been introduced to the APO [Australian Post Office] .... The impact of these developments has resulted in the diversion of significant ADP and user resources to plan and design systems to produce outputs originating from the use of this equipment. [Australian Post Office 1971]
Interviewees discussed a range of detailed difficulties with a long-term senior executive commenting that, "We just couldn't construct a model that worked under test File conditions were too complicated and too volatile. In reality most of the systems we were building were at least partly out of date before they hit production. COMIS testing showed me that integrated systems weren't on." This reference to data issues was echoed in other interviews; there was strong agreement that despite the planners' recognition of data issues, the most intractable practical difficulties in synchronizing the two initial systems developments were those of aligning ftle structures and conditions (Australian Post Office 1971 ). Interviewees reported a range of problems stemming particularly from difficulties in developing data structures capable of supporting multi-stage processes comprising a number of distinct functions.
The review did not formally signify the end of CO MIS, but references to the integrated system and its reengineering intents were quietly dropped. A subsequent major strategic planning document (Telecom Australia 1977) contained no reference to COMIS.
5. INTERPRETATION OF THE CASE STUDY FINDINGS
With a further twenty years of IS theory and research as a foundation for 20/20 hindsight, it can be concluded that COMIS was an extremely optimistic plan which had virtually no chance of success.
98 Part Two Images of Practice: Business Process Reengineering
While CO MIS was specified in concrete tenns, and at a level of detail sufficient to provide a basis for the development of implementation plans, the study showed that a number of the key issues were discussed solely in abstract terms and without specific consideration of the practical issues involved. This was particularly so first in relation to the view that Telecom's organizational structures were unduly rigid, compartmentalized and suitable for redesign and second in the discussion of data design and management issues.
5.1 Computing Technology and Business Process Redesign
Interviewees confirmed the view that CO MIS was intended to provide a blueprint for a series of steps in which computing technology would enable the restructuring of what were seen to be, particularly in the engineering areas, the excessively bureaucratic processes associated with functions such as billing, service provisioning, service restoration, stores management and network planning. One of the striking things about the study group's report is that the planners did not feel the need to qualify any of their remarks in these respects. An interviewee commented that the preparation of CO MIS was an unusually harmonious activity for Telecom, and that this reflected a strong sense in the technical arms of the organization that the advent of computing had fundamentally "changed the rules."
Weizenbaum has commented on the tendency for the power of new technology to "blind" people to its limitations. In our interpretation, the technological bias within Telecom's organizational culture was a significant factor which predisposed planners to believe in the efficacy of technology-mediated solutions to business problems generally. Interviewees commented to the effect that this predisposition was being reinforced during this period by the awareness in technical areas of the world-wide explosion of interest in computing as a business technology (discussed in Gibson and Nolan 1974). We infer that, from a technological perspective, the potential for the application of computers to alleviate existing and anticipated business problems was perceived to be so great as to preclude a critical evaluation of the full range of issues involved in reorganizing process structures.
The result in our view was an implicit but powerful assumption that a computing-mediated restructure of existing inefficient business processes was both desirable and fundamentally unproblematic. The exclusively technical focus identitied earlier suggests both a lack of awareness of, and perhaps a lack of interest in, the true complexity of the business issues involved. In the fmal analysis, the result was a highly detailed and specific plan designed to solve a number of "engineering" problems perceived to exist in organizational structures on the basis of a superficial and very high-level analysis.
5.2 Difficulties in Implementing COMIS
The case discussion highlighted the problems encountered in attempting to implement CO MIS. In retrospect, most of those now look to have been inevitable. It is, however, of value to consider some specific issues in more detail as follows:
I. Managing business support requirements in an evolving environment. 2. Planning and managing corporate data.
IT-mediated business reengineering in Telecom Australia 99
Managing Business Support Requirements in an Evolving Environment. The COMIS Review report noted that the difficulties of building and implementing large systems had been underestimated (Australian Post Office 1971 ). As already discussed, the plan for automating the set of business processes was treated as essentially an engineering issue, and not a particularly difficult one at that. Despite this, it is difficult to criticize the planners for what has only since been recognized as an ongoing problem for the IS industry. Even now, the problems of managing and stabilizing user requirements in a dynamic business environment remain a major issue for systems analysts and designers (for instance, Curtis et al. 1988; Wastell and Newman 1993; Brooks 1987).
A point made in the COMIS Review, and strongly supported by interviewees, was that the technology itself was a major contributor to the environmental volatility which proved so difficult to manage. The introduction of computers into the business and into the telecommunications network led to almost constant rearrangements of business processes such as the collection of network usage details for billing purposes. Benjamin and Blunt ( 1992) suggest a general recursive relationship between IT and business processes when they point out that
the technology is allowing us to build ever larger and more complex systems, and supporting interdependent business processes will require those larger and more complex systems. Thus IT will continue to be involved in a change process that at the same time it makes more complex.
Although as discussed above the planners were well aware that changes, including the widespread introduction of computing, would occur in the business environment, they evidently failed to anticipate the extent of the effects. The discussion in the CO MIS Review report (Australian Post Office 1971) on staffing and resource impacts makes it clear that there were continual conflicts between requirements based on emerging needs to apply computing technology in a specifically business support role and those based on the wish to apply it in the interests of enabling process integration and restructure. The result in effect was a dynamic, constantly moving target for systems development. A concluding comment in one interview was that, "Finally we had to face the fact that the systems we were developing were out of date before they hit production. Things were even worse when we tried to match systems up."
Planning and Managing Corporate Data. Broadly speaking, the data concepts in CO MIS were consistent with the philosophy underpinning information engineering. This has been defined as the view that systems should be designed "with the total organization's information needs in mind and based on an assumption that a relatively stable group of data entities lies at the center of an organization's information processing needs" (Goodhue et al. 1992). In recent years, this idea has been used to ground the development of a number of planning methodologies (for instance, Martin 1982, 1990).
In our view, the data problems experienced in attempting to implement CO MIS are a pmticularly valuable illustration of the risks involved in assessing specific practical issues at an abstract level. When considering data requirements such as standardized code-sets, corporate data definitions and common data structures, the planners were in a sense unfortunate to have few if any precedents on which to base an assessment of possible problems and inhibitors. Extensive
100 Part Two Images of Practice: Business Process Reengineering
subsequent research indicates that similar problems to those encountered with CO MIS have been experienced by almost all finns conunitting to corporate data planning (for instance, in Allen and Boynton 1991; Goodhue et al. 1992; Earll993). A common finding from these studies is that very careful and detailed planning and management is required even for limited progress to be made in developing implementable corporate data structures.
The level of abstraction in the COMIS analysis was such that data consistency was viewed simply as a necessary precondition for the development and implementation of an integrated structure, and therefore as something which had to occur. In effect, issues of a fundamental nature from a practical perspective remained below the planners' threshold of attention. Yet in the event the problems of defming generally usable structures, reconciling conflicting views and dealing with the dynamics of business change proved insuperable even in the context of the first two applications.
6. THE IMPLICATIONS OF COMIS FOR BUSINESS REENGINEERING THEORY
We believe that Telecom's experience with COM IS suggests the desirability of critically examining the assumptions underpinning any concept of IT-mediated organizational change.
In our understanding, reengineering theory in its current formulation embeds an equivalent broad assumption to that underpinning CO MIS; i.e., that the effects of IT on the business environment have "changed the rules" of business organization in fundamental ways. The leading claim for the desirability of reengineering is that current business processes are inefticient or obsolete because they are based on principles of organization which became relevant during the industrial revolution but are no longer applicable due primarily to the effects of modem IT (Hammer 1990; Hammer and Champy 1993; Davenport and Short 1990; Venkatraman 1991, 1994). One problem is that it is difficult to discern what the existing principles of organization are being taken to be. Venkatraman, for instance, states that
the current business processes subscribe to a set of organizational principles that responded to the industrial revolution. Organizational concepts such as centralization versus decentralization, span of control, line versus staff, functional specialization, authority-responsibility balance, and administrative mechanisms for coordination and control are all derived from the general principles. Although these concepts are still valid, IT functionality can significantly alter some of these "first principles" of business process redesign. [Venkataraman 1994]
A difficulty is that neither here nor in context are these first principles fully explicated, nor is it clear why the derived concepts would remain valid if the principles change, though a related comment to the effect that IT can "shrink the effects of time and space" (Venkatraman 1994) suggests perhaps that it is the organization of process support structures according to the
principle of the division of labor which is in doubt.
Hammer and Champy in their tum identify IT as a "powerful solution" and recommend that companies should "seek the problems it might solve." While this seems to be good general
IT-mediated business reengineering in Telecom Australia 101
advice, it does not present as a rigorous justification for the claims made for reengineering as a specific theory. At a more detailed level of discussion they adopt a "broad brush" approach similar in principle to that of Davenport (1993) and identify a number of IT capabilities with specific relevance to process redesign. Much of this material is based on the view that technology can be used to disrupt inappropriate "rules" of organizational behavior and a number of largely independent examples of rule transformations are presented. The connecting link again appears to be that IT has reduced the relevance of specialization and hence of process structures based on different functional domains of skill and expertise. Again, however, there is little or no attempt at formal justification. To take a specific example, one "old rule" is presented as "only experts can perform complex work"; the disruptive technology recommended is expert systems and the new rule as "a generalist can do the work of an expert" (Hammer and Champy 1993). This totally ignores a large body of literature indicating that reliable and organizationally valuable expert systems are still very rare and that in practice the search for, and definition of, expertise is extremely problematic (see Eaves [ l993a, l993b] for an overview of the literature and a discussion of the issues involved).
We take the overall justification for the claim that basic organizational principles have changed to be that IT has developed to a point at which it can be effectively applied in a general integrative role within a large organization. It seems correct to argue that large organizations have traditionally been able to survive only on the basis of task and skill specialization, and we agree with the view that there is usually an increase in overhead staff performing integrative functions as an organization grows larger, a problem which integrated IT may be able to alleviate to an extent (Hammer and Champy 1993). It also seems correct to infer that process design must take into account the full range of possible roles IT can play in process support, up to and including the automation of all steps.
It still seems to us that a leap of faith is required to assume that there is a new basis for process design. Findings from studies into the effects of automation on work practices do demonstrate that skill sets are changing in complex and difficult to manage ways but not that requirements for skill and task difl:erentiation are being reduced (Zuboff 1988). The rate of environmental change is such that the optimization of structures and processes seems inherently unachievable. In a detailed analysis of the issue, Bergquist (1993) comments that the problem of scale is a major issue in the analysis of large organizations. He suggests that process and structural inefficiencies which are "obvious" at a high level cease to be obvious as a finer grain of analysis is undertaken, and the impacts of requirements to change and stabilize new arrangements become apparent.
In our interpretation, COMIS provides a clear illustration of the risks involved in taking a highly abstract view of organizational change. The COMIS analysis was conducted at a level that helped to conceal the real issues of dealing with the application of computers in a dynamic rather than static environment. We suggest that a similarly high-level view of process problems in modem business is to be found in reengineering theory, and that a convincing case has yet to be made for the claim of a change in the principles of process design. While it is obvious that both IT and the business environment have changed dramatically in the years since 1970, it seems by no means equally clear that there has been any qualitative change in the relationship between IT and business processes.
102 Pan Two Images of Practice: Business Process Reengineering
COMIS indicates that issues affecting the role of IT in an organization particularly tend to arise at the nexus of business and functional planning, where long-term structural objectives need to be evaluated in the light of support for shorter term business strategies and requirements, a fmding substantially supported by other research (for instance, Allen and Boynton 1991; Keen 1991). The types of systems and data problems encountered appear to be analogous with those still unresolved in current IS and IT theory. The implication in our view is that the effective application of IT in a large enterprise is a multi-dimensioned activity and that it over-simplifies the situation to conceive the role of IT in relationship to process design as an enabler of change.
7. CONCLUSION AND SUGGESTIONS FOR FURTHER RESEARCH
A tendency has emerged in the reengineering literature to downplay the view that process redesign should be undertaken in the interests of radical change. It ha~. for instance, been suggested by one of the original proponents of recngineering that the movement has already "peaked" and that business process redesign is in the throes of being combined with other process-oriented approaches in an "integrated process management approach" (Davenport and Stoddard 1994). In our view, this effectively begs the question as to the substance in reengineering theory.
A relevant background consideration in our view is that business management is becoming increasingly concerned with IT costs. There appears to be a significant business risk involved when the implications of adopting a business reengineering approach to organizational change arc primarily worked out on the basis of practice. In reviewing the issues involved in the management of IT economics, Keen comments that evaluations of IS function effectiveness in business tend to be generally negative. He deems this unsurprising on the basis that "with costs growing far faster than business profits, evidence of pay-off limited or absent, and small sanctioned increases in annual budgets generating long-term expenses, it i~ no wonder that corporate IS units are under siege." The growing trend toward outsourcing has been interpreted as a reaction to recurring failures by IT units to deliver on the promise of technology, and in this context has been described as "the revenge of the business manager'' (Fitzgerald 1994).
COMIS, in our view, is simply one instance of a past IT failure deriving to a considerable extent from an excessively abstract analysis of the issues involved. Our concern here is that the widespread application of reengineering principles by companies which have failed to appreciate the magnitude of the tasks involved will add to the list of such failures. In this regard, Davenport and Stoddard's comment that "the popular management literature, by relying too much on hype and too little on research, common sense and the lessons of the past. has created more myth than methodology" is indicative of the problems for business. The danger is that companies confronting the problems imposed by the complexity of modern business will be misled by the simplicity of a statement such as "don't automate, obliterate" (Hammer and Champy 1993) as to the likely benefits of reengineering. Such detailed case study literature as is available suggests that, even where reengineering concepts have been applied purely in the interests of incremental process improvement, the costs are high, the learning curve steep and the risks significant (Caron, Jarvenpaa and Stoddard 1994).
IT-mediated business reengineering in Telecom Australia 103
It seems clear, however, that reengineering is here to stay as an approach to planned process redesign. Based on our analysis of CO MIS, we believe that the key research requirement is to build up a body of case studies of sufficient depth to illustrate the full organizational implications of applying reengineering principles in a range of different enterprises and industries. Areas for examination would include the impacts on staff training, skills development and job structures, the direct costs of action, the impacts on existing systems, the need for new systems, data structure requirements and the time frames required to specify, develop and then implement changed business processes. We believe that it is only through detailed research into major reengineering initiatives that an adequate understanding of the validity of the assumptions involved will be developed.
8. REFERENCES
Allen, B. R., and Boynton, A. C. "Information Architecture: In Search of Efficient Flexibility. MIS Quarterly, Volume 15, Number 4, 1991.
Australian Post Office. A.D.P. Project: Exploratory Study Terms of Reference. Melbourne: Australian Post Oftice, 1959.
Australian Post Office. A. D.P. Project: Exploratory Study- Report. Melbourne: Australian Post Office, 1961.
Australian Post Office. Joint Committee of Public Accounts Inquiry into Automatic Data Processing: Submission by Postmaster-General's Department. Melbourne: Australian Post Office, 1966.
Australian Post Office. Report of the Study Group on Application of Computers. Melbourne: Australian Post Office, 1968.
Australian Post Office. COM IS Review. Melbourne: Australian Post Office, 1971.
Benjamin, R. 1., and Blunt, J. "Critical IT Issues: The Next Ten Years." Sloan Management Review, Volume 34, Number I, 1992.
Berquist, W. The Postmodem Organisation: Mastering the Art of Irreversible Change. San Francisco: Josey-Bass Publishers, 1993.
Brooks, F. P. 'The Silver Bullet: Essence and Accidents in Software Engineering." Computer, Aprill987.
Caron, J. R.; Jarvenpaa, S. L.; and Stoddard, D. B. "Business Reengineering at CIGNA Corporation: Experiences and Lessons Learned from the First Five Years." MIS Quarterly, Volume 18, Number 3, 1994.
Curtis, B.; Krasner, H.; and Iscoe, N. "A Field Study of the Software Design Process for Large Systems." Communications of the ACM, Volume 31, Number 11, 1988.
104 Part Two Images of Practice: Business Process Reengineering
Davenport, T. H. Process Work: Reengineering Work Through Information Technology. Boston: Harvard Business School Press, 1993.
Davenport, T. H., and Short, J. E. 'The New Industrial Engineering: Infonnation Technology and Business Process Redesign." Sloan Management Review, Volume 32, Number I, 1990.
Davenport, T. H., and Stoddard, D. B. "Reengineering: Business Change of Mythic Proportions." MIS Quarterly, Volume 18, Number 2, 1994.
Earl, M. J. "Experiences in Strategic Infonnation Systems Planning." MIS Quarterly, Volume 17, Number I, 1993.
Earl, M. J. "The New and the Old of Business Process Redesign." Journal of Strategic Information Systems, Volume 3, Number I, 1994.
Eaves, D. "Experts, Expertise and Professionals' Explanations." Monash University Department oflnfonnation Systems Working Paper Series, Number 19, 1993a.
Eaves, D. "Professionals, Professions and the Curiosities of Clinical Practice." Monash University Department of Infonnation Systems Working Paper Series, Number 20, 1993b.
Fitzgerald, G. "Outsourcing of IT in the United Kingdom: A Legitimate Strategic Option?" Proceedings of the Fifth Australian Conference on Information Systems. Melbourne: Monash University, 1994.
Firestone, W. A. "Accommodation: Towards a Paradigm-Praxis Dialectic." In E. G. Guba (Editor), the Paradigm Dialog. Newbury Park, California: Sage Publications, 1990.
Geertz, C. The Interpretation of Cultures. New York: Basic Books, 1973.
Gibson, C. F., and Nolan, R. L. "Managing the Four Stages of EDP Growth." Harvard Business Review, January-February 1974.
Goodhue, D. L.; Kirsch, L. J.; Quillard, J. A.; and Wybo, M. D. "Strategic Data Planning: Lessons from the Field." MIS Quarterly, Volume 16, Number I, 1992.
Hammer, M. Reengineering Work: Don't Automate, Obliterate." Harvard Business Review, Volume 68, Number 4, 1990.
Hammer, M., and Charnpy, J. Re-engineering the Corporation: A Manifesto for Change. London: Nicholas Brealey Publishing, 1993.
Keen, P. G. W. Shaping the Future: Business Design Through Information Technology. Cambridge: Harvard Business School Press, 1991.
IT-mediated business reengineering in Telecom Australia 105
Kettinger, W. J.; Grover, V.; and Segars, A. H. "Do Strategic Systems Really Pay Off?" Information Systems Management, Winter 1995.
Martin, J. Information Engineering: Planning and Analysis. Englewood Cliffs, New Jersey: Prentice-Hall, 1990.
Martin, J. Strategic Data-Planning Methodologies. Englewood Cliffs, New Jersey: PrenticeHall, 1982.
Moyal, A. Clear Across Australia: A History of Telecommunications. Melbourne: Thoma~ Nelson Australia, 1984/
Simon, H. A. "The Shape of Automation." In Z. W. Pylyshyn (Editor), Perspectives on the Computer Revolution. Englewood Cliffs, New Jersey: Prentice-Hall, 1970.
Telecom Australia. Information Systems Strategic Plan 1977-1987. Melbourne: Telecom Australia, 1977.
Venkatrarnan, N. "IT-Enabled Business Transformation: From Automation to Business Scope Redefinition." Sloan Management Review, Winter 1994, pp. 73-87.
Venkatrarnan, N. "IT-Induced Business Reconfiguration." In M. S. Scott Morton (Editor), The Corporation of the 1990s. Oxford: Oxford University Press, 1991.
Wastell, D., and Newman, M. "The Behavioral Dynamics of Information System Development: A Stress Perspective." Accounting, Management and Information Technology, Volume 3, Number 2, 1993.
Weizenbaum, J. Computing Power and Human Reason: From Judgment to Calculation. San Francisco: W. H. Freeman and Company, 1976.
Yates, J., and Benjamin, R.I. 'The Past and Present as a Window on the Futur." In M. S. Scott Morton (Editor), The Corporation of the 1990s. Oxford: Oxford University Press, 1991.
Zuboff, S. in the Age of the Smart Machine: The Future of Work and Power. New York: Basic Books, 1988.
About the Authors
Doug Hamilton spent twenty-seven years in industry as an IS professional. He worked as a programmer, systems analyst, systems designer and project manager. During the 1980s and early 1990s, he was involved in IS strategic planning in Telecom Australia. Executive assignments have included management of the applications integration function, coordination and management of IS strategic plan implementation, management of the corporate MIS and EIS initiatives and management of the internal business consultancy function. He has been engaged in higher degree
106 Part Two Images of Practice: Business Process Reengineering
studies at Monash University since 1993 and is now enrolled as a doctoral student in the Information Systems Department. Mr. Hamilton's primary research interest currently is strategic
information systems planning.
Martin Atchison has worked variously as an engineer, town planner, and systems analyst.
Currently employed as a lecturer in Information Systems at Monash University, his primary
research interests are in information systems strategic planning and geographical information systems.
